Vehicle defogging methods, vehicles, storage media and computer program products

Abstract

This application discloses a vehicle defogging method, a vehicle, a storage medium, and a computer program product, relating to the field of vehicle technology. This application calculates the scene confidence of a predefined fogging scene based on vehicle visual data and preset prompts related to the predefined fogging scene, eliminating reliance on maps / fences and achieving accurate identification of fogging scenes in unfamiliar environments. Furthermore, it constructs a total condensation risk index by combining scene confidence, in-vehicle climate data, and vehicle operating condition data. When the total condensation risk index exceeds a first preset threshold, pre-defogging processing is performed before fogging occurs. Multi-source data fusion improves the accuracy of condensation risk prediction, avoiding prediction errors based on a single dew point temperature difference, thereby reducing the fogging rate.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and in particular to a vehicle defogging method, a vehicle, a storage medium, and a computer program product. Background Technology

[0002] With the development of the automotive industry, cars are increasingly integrated into our lives. During vehicle operation, if the outside temperature is low, the windshield is prone to fogging due to the higher temperature and humidity of the air exhaled by passengers, potentially affecting the safety of occupants. Automatic window defroster can reduce driving risks and has become a hot topic in the fields of smart cockpits and autonomous driving safety.

[0003] In related technologies, automatic defogging often relies on pre-triggering of maps / fences, or on automatic defogging based on the difference between dew point temperature and the temperature inside and outside the vehicle. However, the former relies on prior marking and coverage, which is difficult to adapt to unfamiliar environments and is prone to false triggering; the latter has a larger error, and often the defogging function is only triggered after obvious fog appears on the windshield, which affects driving safety.

[0004] Therefore, there is an urgent need for a method that can accurately identify fogging scenarios to reduce false triggering rates and perform pre-fogging treatment in advance. Summary of the Invention

[0005] The main objective of this application is to provide a vehicle defogging method, a vehicle, a storage medium, and a computer program product, which aims to provide a method that can accurately identify fogging scenarios, reduce false triggering rates, and perform pre-defogging treatment in advance.

[0006] To achieve the above objectives, this application proposes a vehicle defogging method, the method comprising: acquiring vehicle visual data, in-vehicle climate data, and vehicle operating condition data; performing semantic fusion analysis based on the vehicle visual data and preset prompt words to obtain the matching probability of the prompt words, wherein the preset prompt words are related to a predefined fogging scenario; obtaining the scenario confidence level of the predefined fogging scenario based on the matching probability of the prompt words; calculating a total condensation risk index based on the scenario confidence level, the in-vehicle climate data, and the vehicle operating condition data; and performing pre-defogging processing if the total condensation risk index is greater than a first preset threshold.

[0007] In one embodiment, calculating the total condensation risk index based on the scene confidence level, the in-vehicle climate data, and the vehicle operating condition data includes: determining a scene gain based on the scene confidence level and a priori amplification parameters corresponding to the predefined fogging scene, provided that the maximum value of the scene confidence level is greater than or equal to a scene confidence level threshold; calculating a dew point safety margin and a short-term temperature / humidity change rate based on the in-vehicle climate data, wherein the dew point safety margin is used to characterize the temperature difference from the condensation critical point; and calculating the total condensation risk index based on the dew point safety margin, the short-term temperature / humidity change rate, the scene gain, and the vehicle operating condition data.

[0008] In this embodiment, by combining the scenario confidence levels of multiple predefined fogging scenarios, in-vehicle climate data, and vehicle operating condition data, the vehicle constructs a scientifically systematic total condensation risk index. This deeply couples physical sensing data with semantic prior information, thus constructing a quantifiable total condensation risk index.

[0009] In one embodiment, the predefined fogging scenario includes an outdoor scenario, an entry scenario, an underground parking scenario, and an exit scenario, and the prior amplification parameter corresponding to the outdoor scenario and the underground parking scenario is smaller than the prior amplification parameter corresponding to the entry scenario or the exit scenario.

[0010] In this embodiment, the intensity of thermal and moisture abrupt changes varies depending on the scenario. For example, the intensity of thermal and moisture abrupt changes is greater in the inbound or outbound scenarios, while it is smaller in the outdoor and underground parking scenarios. Therefore, the prior amplification parameters for the outdoor and underground parking scenarios can be set to be smaller than the prior amplification parameters for the inbound or outbound scenarios.

[0011] In one embodiment, the pre-defogging process performed when the total condensation risk index is greater than a first preset threshold includes: inputting initial defogging parameters, in-vehicle climate data, and vehicle operating condition data into a predictive control model; the predictive control model predicts short-term in-vehicle climate data and optimizes defogging parameters based on a predefined in-vehicle thermo-humid dynamics model to obtain target defogging parameters that meet the constraints; and performing pre-defogging processing based on the target defogging parameters. In this embodiment, a predictive control model is used for dynamic optimization to obtain the target defogging parameters.

[0012] In one embodiment, the target defogging parameters satisfy one or more of the following constraints: maintaining windshield visibility, low energy consumption, and in-vehicle temperature. In this embodiment, the optimization objective of the predictive control model takes into account energy consumption, comfort, and noise under the hard constraint of visibility, and controls the actuator rate of change to suppress oscillations.

[0013] In one embodiment, after performing pre-defogging processing based on the target defogging parameters, the method further includes: receiving defogging parameters manually adjusted by the user; and updating the weight parameters in the predictive control model based on the manually adjusted defogging parameters. This embodiment achieves personalized comfort control by learning user preferences online, gradually aligning the defogging control curve with individual user preferences; significantly reducing driver intervention rates and enhancing vehicle intelligence and user trust.

[0014] In one embodiment, obtaining the scene confidence of the predefined fogging scene based on the matching probability of the prompt words includes: calculating the initial confidence of the predefined fogging scene based on the matching probability of the prompt words; and smoothing the initial confidence based on a preset temporal smoothing coefficient to obtain the scene confidence of the predefined fogging scene.

[0015] In this embodiment, temporal smoothing technology can improve the stability of semantic recognition and solve the recognition jitter problem caused by environmental interference; it can also effectively suppress recognition jitter caused by factors such as changes in ambient lighting and interference from obstructions.

[0016] In one embodiment, the in-vehicle climate data includes: the inner surface temperature of the windshield and the in-vehicle humidity; the inner surface temperature of the windshield is obtained based on the vehicle's windshield temperature sensor, or estimated based on the first law of thermodynamics of the external ambient temperature; the in-vehicle humidity is obtained based on the vehicle's humidity sensor, or estimated based on the acquired vehicle volume, vehicle ventilation mass flow rate, number of occupants and metabolic level, and external ambient humidity. This embodiment provides multiple implementation methods for the inner surface temperature of the windshield and the in-vehicle humidity, applicable to various vehicle models.

[0017] In one embodiment, after performing pre-defogging treatment when the total condensation risk index is greater than a first preset threshold, the method further includes: acquiring the vehicle visual data, in-vehicle climate data, and vehicle operating condition data in real time; updating the total condensation risk index based on the vehicle visual data, the in-vehicle climate data, and the vehicle operating condition data; and stopping the pre-defogging treatment when the total condensation risk index is less than or equal to a second preset threshold, wherein the second preset threshold is less than the first preset threshold.

[0018] This embodiment employs triggering logic with hysteresis, which effectively suppresses frequent control switching and actuator oscillation; improves the smoothness of vehicle operation and user experience, and avoids the discomfort caused by "on / off" control.

[0019] In addition, to achieve the above objectives, this application also proposes a vehicle comprising: a plurality of displays, a speaker, a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the vehicle defogging method described above.

[0020] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the vehicle defogging method described above.

[0021] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the vehicle defogging method described above.

[0022] The vehicle defogging method provided in this application calculates the scene confidence of a predefined fogging scene based on vehicle visual data and preset prompts related to the predefined fogging scene, eliminating reliance on maps / fences and achieving accurate identification of fogging scenes in unfamiliar environments. Furthermore, it constructs a total condensation risk index by combining scene confidence, in-vehicle climate data, and vehicle operating condition data. When the total condensation risk index exceeds a first preset threshold, pre-defogging processing is performed before fogging occurs. Multi-source data fusion improves the accuracy of condensation risk prediction, avoiding prediction errors based on a single dew point temperature difference, thereby reducing the fogging rate. This embodiment can accurately identify fogging scenes, especially in unfamiliar environments, reducing the probability of false triggering and improving the accuracy of condensation risk prediction. Pre-defogging processing can be performed before fogging occurs, thus reducing the fogging rate. Attached Figure Description

[0023] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0024] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a flowchart illustrating an embodiment of the vehicle defogging method of this application.

[0026] Figure 2 This is a schematic diagram of the vehicle defogging method of this application;

[0027] Figure 3 This is a flowchart illustrating an embodiment of the vehicle defogging method of this application.

[0028] Figure 4 This is a flowchart illustrating an embodiment of the vehicle defogging method of this application.

[0029] Figure 5 This is a flowchart illustrating an embodiment of the vehicle defogging method of this application.

[0030] Figure 6 This is a schematic diagram of the vehicle defogging method of this application;

[0031] Figure 7 This is a schematic diagram of the vehicle architecture for implementing the vehicle defogging method of this application;

[0032] Figure 8 This is a flowchart illustrating Embodiment 2 of the vehicle defogging method of this application;

[0033] Figure 9 This is a flowchart illustrating Embodiment 3 of the vehicle defogging method of this application;

[0034] Figure 10 This is a schematic diagram illustrating the interaction between the vehicle system and the cloud system in this application;

[0035] Figure 11 This is a schematic diagram comparing the vehicle defogging method of this application with traditional solutions;

[0036] Figure 12 This is a schematic diagram of the module structure of the vehicle defogging device according to an embodiment of this application;

[0037] Figure 13 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the vehicle defogging method in the embodiments of this application.

[0038] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0039] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0040] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0041] In this embodiment, for ease of description, the following description uses a vehicle, or a mobile terminal, data storage control terminal, PC (Personal Computer) or other terminal connected to the vehicle's electronic control unit as the execution subject.

[0042] Based on the aforementioned vehicle, the overall concept of the vehicle defogging method of this application is proposed.

[0043] To address the above phenomena, this application provides a vehicle defogging method, comprising: acquiring vehicle visual data, in-vehicle climate data, and vehicle operating condition data; performing semantic fusion analysis based on the vehicle visual data and preset prompt words to obtain the matching probability of the prompt words, wherein the preset prompt words are related to a predefined fogging scenario; obtaining the scenario confidence level of the predefined fogging scenario based on the matching probability of the prompt words; calculating a total condensation risk index based on the scenario confidence level, the in-vehicle climate data, and the vehicle operating condition data; and performing pre-defogging processing when the total condensation risk index is greater than a first preset threshold.

[0044] Thus, based on vehicle visual data and preset prompts related to predefined fogging scenarios, the scene confidence level of the predefined fogging scenario is calculated, eliminating reliance on maps / fences and achieving accurate identification of fogging scenarios in unfamiliar environments. Furthermore, a total condensation risk index is constructed by combining scene confidence, in-vehicle climate data, and vehicle operating condition data. When the total condensation risk index exceeds a first preset threshold, pre-fogging defogging is performed. Multi-source data fusion improves the accuracy of condensation risk prediction, avoiding prediction errors based on a single dew point temperature difference, thereby reducing the fogging rate. This embodiment can accurately identify fogging scenarios, especially in unfamiliar environments, reducing the probability of false triggering and improving the accuracy of condensation risk prediction. Pre-fogging defogging can be performed before fogging occurs, thereby reducing the fogging rate.

[0045] Reference Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the vehicle defogging method of this application, which includes steps S10 to S60.

[0046] Step S10: Acquire vehicle visual data, in-vehicle climate data, and vehicle operating condition data.

[0047] Vehicle visual data refers to the external environment data of the vehicle collected by vehicle cameras (including but not limited to front-view cameras and surround-view cameras). Specifically, it can be image frame data obtained by sampling high-resolution scene videos captured by cameras, with a sampling frequency of, for example, 30fps and an image resolution of, for example, not less than 1920×1080 pixels.

[0048] In-vehicle climate data includes relative humidity inside the vehicle. Car interior temperature Temperature of the inner surface of the windshield Specifically, in-vehicle climate data is obtained by collecting measurement data from ambient temperature sensors, humidity sensors, and glass temperature sensors located inside the vehicle via the CAN / LIN bus.

[0049] For vehicles with limited costs that are not equipped with glass temperature sensors, the ambient temperature can be estimated based on the first law of thermodynamics. The first dynamic estimation model based on the first law of thermodynamics is shown below:

[0050]

[0051] in, The thermal inertia constant of glass, Indicates the coupling gain of defogging / heating on the glass. Indicates external heat transfer gain, Command to heat / defog the vehicle's windshield The external ambient temperature.

[0052] The dynamic estimation model shown in formula (1) above fully considers the thermal inertia effect of the glass, the direct heating effect of the defogging vehicle, and the convective heat transfer with the external environment. , and The method is determined by combining offline calibration and online adaptive methods. It can be obtained in real time based on the vehicle. It can obtain real-time weather data from weather software via the vehicle's network interface.

[0053] For economy vehicles without humidity sensors, humidity is estimated based on vehicle volume, ventilation mass flow rate, number of occupants, metabolic level, and external environmental humidity. The absolute humidity inside the vehicle can be estimated using a second dynamic estimation model based on the principle of mass conservation, from which the relative humidity inside the vehicle can be deduced. The second dynamic estimation model based on the principle of mass conservation is shown in the following equation:

[0054]

[0055] in, For vehicle volume, The mass flow rate for vehicle ventilation (related to air volume / internal / external circulation). External absolute humidity, It is an estimate of humidity based on the number of occupants and metabolic levels. for The absolute humidity inside the car at any time for The absolute humidity inside the car at any given time.

[0056] In obtaining Afterwards, according to The relative humidity inside the car was calculated. For specific calculation methods, please refer to relevant technologies, which will not be elaborated on here.

[0057] Vehicle operating data includes vehicle speed , longitudinal slope angle Compared with external ambient light The vehicle speed can be obtained by collecting measurement data from the vehicle's speed and attitude sensors. and longitudinal slope angle The ambient light level is obtained by collecting measurement data from the vehicle's photosensor. .

[0058] It is worth noting that the above in-vehicle climate data and vehicle operating condition data are for illustrative purposes only. In actual applications, some data can be added or removed as needed without affecting the implementation of the plan.

[0059] Step S20: Based on vehicle visual data and preset prompt words, perform semantic fusion analysis to obtain the matching probability of the prompt words. The preset prompt words are related to the predefined fogging scenario.

[0060] Predefined fogging scenarios can be set, under which sudden changes in temperature and humidity can easily cause fogging on the vehicle's windshield, such as parking and exiting a parking space. Optional. To cover more scenarios, predefined fogging scenarios can also include: indoor scenarios and underground parking scenarios.

[0061] The preset prompts are related to predefined fogging scenarios and cover characteristic elements of underground parking garages, including structural features (such as underground parking garage entrances, exits, ramps, and gates), visual identifiers (such as fluorescent lights, ventilation louvers, yellow and black zebra crossings, and concrete pillars), and textual symbols (such as P signs, floor signs, and directional signs), providing multi-dimensional semantic clues.

[0062] The matching probability of the prompt words is obtained through semantic fusion analysis based on vehicle visual data and preset prompt words. For details, please refer to the appendix. Figure 2 Vehicle visual data (obtained by preprocessing real-time video streams) and a set of pre-defined prompts are input into a specially optimized Vision Language Model (VLM) to obtain the matching probability of each prompt. .specific

[0063] Step S30: Based on the matching probability of the prompt words, obtain the scene confidence of the predefined fogging scene.

[0064] It is achievable, as shown in the attached document. Figure 3 As shown, step S30 may specifically include steps S301 to S302.

[0065] Step S301: Calculate the initial confidence level of the predefined fogging scenario based on the matching probability of the prompt words.

[0066] As attached Figure 2 As shown, it can be based on a preset weight matrix. The initial confidence level for the predefined fogging scenario is obtained by aggregating the matching probabilities of the prompt words. , .in, For the first The predicted probability of each prompt word For prompt words up to the 1st Mapping weights for different scenarios.

[0067] Each predefined fogging scenario corresponds to a preset weight matrix. Furthermore, the preset weight matrix varies depending on the scenario. For example, in the preset weight matrix corresponding to the inbound scenario, the weight values ​​of "underground garage entrance" and "gate" are relatively large, while the weight values ​​of "exit" and "ramp" are relatively small; while the opposite is true for the outbound scenario.

[0068] Step S302: Smooth the initial confidence level based on the preset temporal smoothing coefficient to obtain the scene confidence level of the predefined fogging scene.

[0069] Based on preset time smoothing coefficients For the initial confidence level Smoothing is performed to suppress recognition jitter caused by factors such as changes in ambient lighting and interference from obstructions. Specifically, see attached... Figure 2 As shown, this can be achieved by adding timing consistency constraints, as shown in the following formula:

[0070]

[0071] in, For the smoothed first Scene confidence of similar scenarios This is the time series smoothing coefficient.

[0072] This yields scene confidence vectors for multiple predefined fogging scenarios. ,in, For outdoor scenes, scene confidence, For the scene confidence of the data entry scenario, For the scene confidence of the underground parking lot scene, The scenario confidence level for outbound scenarios.

[0073] This embodiment can improve the stability of semantic recognition through temporal smoothing technology, solve the problem of recognition jitter caused by environmental interference, and effectively suppress recognition jitter caused by factors such as changes in ambient lighting and interference from obstructions.

[0074] Step S40: Calculate the total condensation risk index based on scenario confidence, in-vehicle climate data, and vehicle operating condition data.

[0075] In this embodiment, a scientific and systematic total condensation risk index is constructed for the vehicle by combining the scene confidence levels of multiple predefined fogging scenarios, in-vehicle climate data, and vehicle operating condition data. This deeply couples physical sensing data with semantic prior information, thus constructing a quantifiable total condensation risk index.

[0076] In one feasible implementation, such as Figure 4 As shown, step S40 can specifically include steps S401 to S404.

[0077] Step S401: Determine whether the maximum value of the scene confidence score is greater than or equal to the scene confidence score threshold. If yes, proceed to step S402; otherwise, return to step S10.

[0078] First, determine the maximum scene confidence score. The scene confidence scores for multiple predefined fogging scenes are: , , , Take the maximum value of the scene confidence score, and then compare the maximum value of the scene confidence score with the scene confidence score threshold. Comparison, The maximum value of the scene confidence score is greater than or equal to the scene confidence score threshold. In this case, further calculations are performed to ensure the reliability of the recognition results, as shown in the attached document. Figure 2 As shown. The confidence gating indicator is shown in the following formula:

[0079]

[0080] in, For confidence gating indication, when When the value is 1, execute step S402; when When the value is 0, return to step S10.

[0081] It is achievable, when the confidence level is... Below the scene confidence threshold Alternatively, if a critical sensor malfunctions or optimization fails, the timing smoothing in step S302 will be replaced with a conservative regular curve (fixed defogging / external circulation and medium-high airflow) to expand the safety margin and avoid unexplained actions.

[0082] Step S402: Determine the scene gain based on the scene confidence and the prior amplification parameters of the corresponding predefined fogging scene.

[0083] To map semantic transition confidence to the amplification gain of thermal-humidity transitions, scene confidence based on multiple predefined fogging scenarios is used. , , , The scene gain is calculated by taking into account the prior amplification parameters for the corresponding predefined fogging scenario. This scene gain enables the vehicle to predict environmental abrupt changes based on visual semantic information before physical sensors detect significant changes.

[0084] Assume the prior amplification parameter for the outdoor scene is The prior amplification parameter corresponding to the warehousing scenario is: The prior magnification parameters corresponding to the scene inside the basement are: The prior amplification parameter corresponding to the outbound scenario is: Then scene gain It can be expressed by the following formula:

[0085]

[0086] It is feasible to determine the intensity of thermal and moisture abrupt changes in different scenarios. For example, the intensity of thermal and moisture abrupt changes is greater in inbound or outbound scenarios, while it is smaller in outdoor and underground parking scenarios. Therefore, the prior amplification parameters for outdoor and underground parking scenarios can be set to be smaller than the prior amplification parameters for inbound or outbound scenarios.

[0087] Step S403: Calculate the dew point safety margin and the short-term temperature / humidity change rate based on the in-vehicle climate data. The dew point safety margin is used to characterize the temperature difference from the critical point of condensation.

[0088] In-vehicle climate data includes in-vehicle temperature Relative humidity inside the car and the temperature of the inner surface of the windshield The dew point safety margin (DPM) is calculated based on in-vehicle climate data. The dew point safety margin characterizes the temperature difference from the critical point of condensation.

[0089] DPM

[0090] in, This indicates the dew point temperature, which is determined based on the interior temperature of the vehicle. and relative humidity inside the car Calculated.

[0091]

[0092]

[0093] in, ℃ is a commonly used Magnus approximation parameter.

[0094] It is achievable, in At 3℃, the vehicle is in a safe state; when the DPM drops to the 1-3℃ range, the vehicle enters a warning state; when At 1℃, there is a high risk that the vehicle will soon fog up.

[0095] To accurately characterize the short-term temperature and humidity changes caused by the transition between entering and leaving the warehouse, time-series differential analysis was performed on the temperature and humidity signals. This represents the short-term rate of temperature change. This represents the short-term humidity change rate.

[0096]

[0097]

[0098] In this embodiment, the short-term temperature / humidity change rate is an important leading indicator for predicting impending environmental abrupt changes.

[0099] Step S404: Calculate the total condensation risk index based on dew point safety margin, short-term temperature / humidity change rate, scene gain, and vehicle operating condition data.

[0100] In this embodiment, the dew point safety margin is combined with the temperature difference characterizing the distance from the dew point critical point. Short-term temperature change rate is an important leading indicator for predicting impending environmental upheavals. and short-term humidity change rate ; Scene gain used to predict environmental change trends in advance based on visual semantic information before physical sensors detect significant changes. Vehicle operating condition data characterizing vehicle status: vehicle speed , longitudinal slope angle Compared with external ambient light Construct a total condensation risk index. This unifies physical quantities and semantic priors to the same controllable scale. Total condensation risk index It can be represented by the following formula:

[0101]

[0102] in, Indicates the degree to which the area has entered a danger zone; Reflects the intensity of sudden changes in the thermal environment; Consider the role of increased humidity in promoting fogging; Provides predictive prior information; This indicates that relative wind speed at high speeds helps prevent fogging; This indicates that steep ramps typically correspond to the entrance and exit of underground parking garages; As a low-light indicator, the lighting conditions inside the basement are significantly different. Low illumination threshold; The Sigmoid mapping maps the risk index to the [0,1] interval; These are the parameters obtained by fitting offline / online data.

[0103] It is understandable that when DPM decreases (glass temperature approaches / falls below dew point), the rate of temperature / humidity change increases, or the gain in inbound / outbound scenarios increases, This will significantly increase, indicating that "the risk of fog is imminent."

[0104] Step S50: Determine whether the total condensation risk index is greater than the first preset threshold. If yes, proceed to step S60; otherwise, return to step S10.

[0105] Set the first preset threshold. ,when If the condition indicates a high risk of condensation, step S60 can be executed to perform pre-demisting treatment; otherwise, return to step S10 to perform the next round of total condensation risk analysis.

[0106] It is worth noting that steps S10 to S50 can be executed in real time, and if the total condensation risk index at the current moment is detected to be greater than the first preset threshold, pre-defogging treatment can be performed in a timely manner, thereby intervening in time before the vehicle fogs up.

[0107] Step S60: Perform pre-demisting treatment.

[0108] It is feasible. Vehicles can undergo pre-defogging treatment based on preset defogging parameters, and the risk level can be determined based on the magnitude of the total condensation risk index. Different defogging parameters can be preset for different risk levels, so that when the vehicle is in different condensation risk levels, different defogging parameters can be used for targeted pre-defogging treatment, further reducing the vehicle fogging rate.

[0109] In one feasible implementation, a Model Predictive Control (MPC) model is used for dynamic optimization to obtain the target defogging parameters. Specifically, such as... Figure 5 As shown, step S60 above includes steps S601 to S603.

[0110] Step S601: Input the initial defogging parameters, in-vehicle climate data, and vehicle operating condition data into the predictive control model.

[0111] Step S602: The predictive control model predicts short-term in-vehicle climate data and optimizes defogging parameters based on a predefined in-vehicle thermo-humid dynamics model, and solves for the target defogging parameters that meet the constraints.

[0112] As attached Figure 6 As shown, in this embodiment, the initial defogging parameters are the parameters of the heating, ventilation, and air conditioning (HVAC) system of the vehicle, including the blower airflow. air supply temperature Compressor start / stop / speed Internal and external circulation settings With windshield heating / defogger Initial defogging parameters can be obtained by acquiring the vehicle's HVAC status parameters.

[0113] The initial defogging parameters, in-vehicle climate data, and vehicle operating condition data are input into the predictive control model. The predictive control model predicts short-term in-vehicle climate data and optimizes defogging parameters based on a predefined in-vehicle thermo-humid dynamics model, and solves for the target defogging parameters that meet the constraints.

[0114] The predefined in-vehicle thermal and hygrodynamic model can be expressed as:

[0115]

[0116] in, Represents the state vector at time k. Represents the predicted state vector at time k+1. The perturbation vector at time k, The defogging parameters, matrices A, B, and E at time k are obtained offline using a vehicle identification method and support online adaptive updates.

[0117] This includes the dew point safety margin (DPM) determined based on in-vehicle climate data, and the in-vehicle temperature. Relative humidity inside the car Temperature of the inner surface of the windshield ; Including vehicle operating data, such as vehicle speed , longitudinal slope angle Compared with external ambient light And so on, which may also include external ambient temperature. External relative humidity .

[0118] The achievable target defogging parameters satisfy one or more of the following constraints: maintaining windshield visibility, low energy consumption, and in-vehicle temperature.

[0119] In this embodiment, the optimization objective of the predictive control model takes into account energy consumption, comfort, and noise under the hard constraint of visibility, and controls the rate of change of the actuator to suppress oscillations. Its loss function is shown in the following formula:

[0120]

[0121]

[0122]

[0123]

[0124] in, For ReLU; For energy consumption models; Designed for cabin comfort; Used to characterize wind noise / perceived blower noise, taking into account the user's subjective experience; To control the increment. Visibility function. can be Mapping.

[0125]

[0126] in, This is the minimum safety margin; This is the scaling factor.

[0127] To ensure the stability of the front windshield camera's perception, a higher threshold is set for the ADAS clear zone. As a hard constraint, it should be satisfied first.

[0128] It is worth noting that the predictive control model MPC solves for the target defogging parameter sequence that satisfies the constraints (i.e., the loss function) in each control cycle, but only uses the first defogging parameter u0 as the target defogging parameter. The next control cycle is then used to re-predict and optimize in order to adapt to real-time disturbances and scene changes.

[0129] Step S603: Perform pre-demisting treatment based on the target demisting parameters.

[0130] Assuming the target demisting parameters predicted by the predictive control model MPC include the blower airflow... air supply temperature Compressor start / stop / speed Internal and external circulation settings With windshield heating / defogger These target demisting parameters are translated into actual actuator commands through precise nonlinear mapping: blower airflow. Converted to blower PWM duty cycle mapping (considering motor characteristics and duct resistance), air supply temperature Convert to mixing damper / temperature damper opening (to achieve precise temperature regulation), compressor start / stop / speed Convert to compressor commands (speed control of variable displacement compressor or PWM control of fixed displacement compressor), internal / external circulation settings. Switch to internal / external circulation mode (precise damper positioning control), windshield heating / defogging. Convert to windshield heating power (power adjustment of resistance heating wire).

[0131] It is worth noting that a local priority strategy is defined in the ADAS clear area. By adjusting the airflow distribution of the defogging vent, a stronger local drying / heating airflow is formed in the camera's field of view, ensuring that the imaging quality of the sensing module is not constrained by the overall airflow balance.

[0132] It is feasible; vehicles can select different weight parameters for predictive control models based on the scene confidence of predefined fogging scenarios. , , , , For example, in the case of warehousing, visibility and defogging response weights can be improved. To ensure driving safety during the transition process; in the underground parking scenario, to emphasize the balance between energy consumption and noise, and to avoid long-term high-power operation; in the exit scenario, to suppress temperature difference shocks to avoid re-fogging, and to prevent reverse fogging when moving from the low-temperature underground parking to the high-temperature outside environment.

[0133] It is achievable; the vehicle has a robust safety rollback mechanism. When VLM semantic recognition confidence is low... Continuously below the scene confidence threshold When key sensors (temperature, humidity) malfunction or have abnormal readings, MPC optimization fails or convergence is difficult, vehicle system computing resources are insufficient or communication fails, etc., the system automatically switches to conservative defogging parameters for pre-defogging. Although this may sacrifice some comfort and energy efficiency, it can ensure basic defogging function and driving safety.

[0134] The vehicle defogging method in the above embodiments can be implemented by a vehicle, which includes, for example, vehicles such as Figure 7 The diagram shows the perception layer, processing layer, control layer, and execution layer.

[0135] Based on the first embodiment of this application, a second embodiment of this application is proposed herein. In this second embodiment, content that is the same as or similar to the embodiments described above can be referred to the above description and will not be repeated hereafter. Based on this, as... Figure 8 As shown, after step S603 above, the vehicle defogging method of this application may further include steps A10 to A20.

[0136] Step A10: Receive the defogging parameters manually adjusted by the user.

[0137] Step A20: Update the weight parameters in the predictive control model based on the manually adjusted defogging parameters.

[0138] After pre-defogging based on the target defogging parameters, if the vehicle receives manually adjusted defogging parameters from the user, these manually adjusted parameters (such as temporarily increasing airflow or preferring quiet operation) can be considered as preference signals and applied to the weight parameters of the predictive control model. , , , , Online fine-tuning allows the control curve to gradually align with individual user sensations, making the output of the predictive control model more aligned with user preferences.

[0139] Weight updates can be performed using vector updates with a forgetting factor, as shown in the following formula:

[0140]

[0141] Among them, w integration , , , , etc.; δu represents the deviation of the control quantity from manual intervention; ρ represents the weight suggestions derived from the association between the offset and the scene; ρ is the learning rate.

[0142] In this embodiment, by learning user preferences online, the defogging control curve gradually matches the individual user's sensory preferences, achieving personalized comfort control; significantly reducing driver intervention rate and improving vehicle intelligence and user trust.

[0143] To enhance user acceptance and long-term trust, an explainable and personalized mechanism can be added. The explainable module compresses complex, multi-dimensional triggering factors such as vehicle visual data, in-vehicle climate data, and vehicle operating condition data into user-friendly natural language expressions (e.g., "Enhanced semantics for parking + DPM reduction exceeding limits", "Parking entrance detected, preventative defogging initiated", "Downhill slope and dimming light, advance defogging preparation"), and presents them as optional prompts on the vehicle's display screen to help users understand the vehicle's proactive intervention logic.

[0144] Based on the first embodiment of this application, a third embodiment of this application is proposed herein. In this third embodiment, content that is the same as or similar to the embodiments described above can be referred to the above description and will not be repeated hereafter. Based on this, as... Figure 9As shown, after step S60 above, the vehicle defogging method of this application may further include steps B10 to B40.

[0145] Step B10: Acquire vehicle visual data, in-vehicle climate data, and vehicle operating condition data in real time.

[0146] Step B20: Update the total condensation risk index based on vehicle visual data, in-vehicle climate data, and vehicle operating condition data.

[0147] Step B30: If the total condensation risk index is less than or equal to the second preset threshold, stop the pre-demisting treatment. The second preset threshold is less than the first preset threshold.

[0148] The vehicle continues to operate, acquiring the latest vehicle visual data, in-vehicle climate data, and vehicle operating condition data in real time. Based on the latest data, the total condensation risk index R(t) is updated. When the updated... Exit pre-control at this time. Here The first preset threshold is set. The hysteresis width is used to suppress trigger jitter and frequent switching. This is the second preset threshold.

[0149] This embodiment employs triggering logic with hysteresis, which effectively suppresses frequent control switching and actuator oscillation; improves the smoothness of vehicle operation and user experience, and avoids the discomfort caused by "on / off" control.

[0150] To achieve sustainable generalization capabilities and engineering implementation, a data closed loop is constructed: the vehicle records multiple raw quantities (time series of physical quantities such as temperature, humidity, pressure, light, and motion), VLM semantic recognition results (confidence scores and feature extraction results for each scenario), DPM and R(t) time series, MPC reference and actuator response, as well as trigger / rollback events, forming a weakly labeled operation log.

[0151] like Figure 10 As shown, these data are uploaded to the cloud after being anonymized for privacy. The cloud-based semantic data factory mines and labels difficult examples with weak supervision, and jointly trains VLM, risk mapping, and control hyperparameters. The resulting model is then distilled and quantized before being distributed to the vehicle, realizing a flywheel update of "online calibration - backhaul - retraining - redeployment".

[0152] Deploying a dedicated semantic data analysis factory in the cloud, it possesses core capabilities such as hard example mining (automatically identifying difficult samples that fail VLM recognition or have abnormal confidence), weakly supervised annotation (automatically annotating by combining vehicle trajectory, map information, etc.), joint optimization training (end-to-end joint training of VLM model, risk mapping function, and MPC control parameters), and model compression and quantization (knowledge distillation and quantization of the trained large model to adapt it to automotive-grade computing platforms).

[0153] In addition, the vehicle offers robust variants with different sensing configurations: when humidity or glass temperature sensors are missing, moisture content approximation and glass thermal model estimation are enabled respectively; when computing resources are limited, a distilled lightweight semantic detector is used to replace the complete VLM, but the interface and control logic remain unchanged, ensuring mass production portability and cost control.

[0154] The vehicle supports sensor configurations at different cost levels, including high-end configuration (glass temperature sensor + high-precision humidity sensor + complete VLM), standard configuration (humidity sensor + thermodynamically estimated glass temperature + complete VLM), and economic configuration (moisture content estimation + thermodynamic estimation + lightweight semantic detector).

[0155] The lightweight mode employs a knowledge-distilled lightweight VLM (model size compressed by more than 60%), a simplified MPC prediction model (reducing state dimensions and prediction time domain length), and an approximate optimization algorithm (reducing real-time computational complexity). Although the recognition accuracy is slightly reduced (3~5%), the core functions remain intact.

[0156] In summary, such as Figure 11 As shown, in this embodiment of the application, the process involves "VLM semantic transition identification → dew point and safety margin physical modeling → condensation risk". The continuous workflow of "feedforward evaluation → MPC rolling optimization under visibility hard constraints → execution / explainability / personalization / safety rollback → cloud-vehicle data closed-loop evolution" upgrades "passive defogging after fogging" to "active prediction before fogging." This significantly reduces the probability of initial fogging and the time required for defogging in scenarios with sudden temperature and humidity changes, such as entering / exiting underground parking garages. It also reduces energy consumption and noise caused by ineffective high-airflow operation, while enhancing user trust and long-term acceptance through personalization and explainability mechanisms. Extensive real-vehicle testing in various typical scenarios has verified that this technology achieves a fogging prevention success rate of over 95% in underground parking garage entry and exit scenarios compared to traditional defogging solutions. It also offers a response time 15-30 seconds faster than passive sensor solutions, reduces average energy consumption by 25-30% through precise control, and user subjective evaluations show that 88% of users believe the experience is significantly improved. Therefore, this constitutes a semantically driven in-vehicle microclimate prediction and defogging control technology system with engineering feasibility and practical application.

[0157] This application also provides a vehicle defogging device, please refer to... Figure 12 The device includes:

[0158] The parameter acquisition module 10 is used to acquire vehicle visual data, in-vehicle climate data, and vehicle operating condition data.

[0159] The semantic analysis module 20 is used to perform semantic fusion analysis based on the vehicle visual data and preset prompt words to obtain the matching probability of the prompt words, wherein the preset prompt words are related to a predefined fogging scenario.

[0160] The confidence calculation module 30 is used to obtain the scene confidence of the predefined fogging scene based on the matching probability of the prompt words.

[0161] The risk index calculation module 40 is used to calculate the total condensation risk index based on the scenario confidence level, the in-vehicle climate data, and the vehicle operating condition data.

[0162] The pre-demisting control module 50 is used to perform pre-demisting treatment when the total condensation risk index is greater than a first preset threshold.

[0163] In practice, the aforementioned risk index calculation module 40 is further configured to: determine a scene gain based on the scene confidence level and the prior amplification parameters corresponding to the predefined fogging scene, when the maximum value of the scene confidence level is greater than or equal to the scene confidence level threshold; calculate a dew point safety margin and a short-term temperature / humidity change rate based on the in-vehicle climate data, wherein the dew point safety margin is used to characterize the temperature difference from the condensation critical point; and calculate a total condensation risk index based on the dew point safety margin, the short-term temperature / humidity change rate, the scene gain, and the vehicle operating condition data.

[0164] It is achievable that the above-mentioned predefined fogging scenarios include outdoor scenarios, inbound scenarios, underground parking scenarios, and outbound scenarios, and the prior amplification parameters corresponding to the outdoor scenarios and underground parking scenarios are smaller than the prior amplification parameters corresponding to the inbound scenarios or the outbound scenarios.

[0165] The aforementioned pre-defogging control module 50 is also used to input the initial defogging parameters, the in-vehicle climate data, and the vehicle operating condition data into the predictive control model; the predictive control model performs short-term in-vehicle climate data prediction and defogging parameter optimization based on a predefined in-vehicle thermo-humid dynamics model, and solves to obtain the target defogging parameters that meet the constraints; and performs pre-defogging processing based on the target defogging parameters.

[0166] It is achievable that the target defogging parameters satisfy one or more of the following constraints: maintaining windshield visibility, low energy consumption, and in-vehicle temperature.

[0167] The vehicle defogging device also includes a feedback control module 60. After pre-defogging based on the target defogging parameters, the feedback control module 60 receives the defogging parameters manually adjusted by the user and updates the weight parameters in the predictive control model based on the manually adjusted defogging parameters.

[0168] The confidence calculation module 30 is also used to calculate the initial confidence of the predefined fogging scene based on the matching probability of the prompt word; and to smooth the initial confidence based on a preset temporal smoothing coefficient to obtain the scene confidence of the predefined fogging scene.

[0169] The in-vehicle climate data includes: the inner surface temperature of the windshield and the in-vehicle humidity; the inner surface temperature of the windshield is obtained based on the vehicle's windshield temperature sensor, or estimated based on the first law of thermodynamics of the external environment temperature; the in-vehicle humidity is obtained based on the vehicle's humidity sensor, or estimated based on the acquired vehicle volume, vehicle ventilation mass flow rate, number of occupants and metabolic level, and external environment humidity.

[0170] The parameter acquisition module 10 is configured to acquire the vehicle visual data, in-vehicle climate data, and vehicle operating condition data in real time after performing pre-defogging treatment when the total condensation risk index is greater than a first preset threshold. The risk index calculation module 40 is configured to update the total condensation risk index based on the vehicle visual data, the in-vehicle climate data, and the vehicle operating condition data. The pre-defogging control module 50 is configured to stop the pre-defogging treatment when the total condensation risk index is less than or equal to a second preset threshold, wherein the second preset threshold is less than the first preset threshold.

[0171] The vehicle defogging device provided in this application, employing the vehicle defogging method described in the above embodiments, can solve the technical problem in related technologies where drivers cannot intuitively perceive the presence and movement trends of surrounding vehicles. Compared with the prior art, the beneficial effects of the vehicle defogging device provided in this application are the same as those of the vehicle defogging method provided in the above embodiments, and other technical features of the vehicle defogging device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0172] This application provides a vehicle, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the vehicle defogging method of the above embodiment 1.

[0173] like Figure 13As shown, the vehicle may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory 1002 or a program loaded from a storage device 1003 into a random access memory 1004. The random access memory 1004 also stores various programs and data required for vehicle operation. The processing unit 1001, the read-only memory 1002, and the random access memory 1004 are interconnected via a bus 1005. An input / output interface 1006 is also connected to the bus. Typically, the following vehicles can be connected to the input / output interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. The communication device 1009 allows the vehicle to communicate wirelessly or wiredly with other devices to exchange data. Although the diagram shows vehicles with various configurations, it should be understood that it is not required to implement or possess all of the vehicles shown. Alternatively, more or fewer vehicles may be implemented or possessed.

[0174] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0175] The vehicle provided in this application, employing the vehicle defogging method described in the above embodiments, can solve the technical problem in related technologies where drivers cannot intuitively perceive the presence and movement trends of surrounding vehicles. Compared with the prior art, the beneficial effects of the vehicle provided in this application are the same as those of the vehicle defogging method provided in the above embodiments, and other technical features of the vehicle are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0176] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0177] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0178] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to perform the vehicle defogging method in the above embodiments.

[0179] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor vehicles or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction-executable vehicle or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0180] The aforementioned computer-readable storage medium may be included in the vehicle or may exist independently and not installed in the vehicle.

[0181] The aforementioned computer-readable storage medium carries one or more programs that, when executed by a vehicle, cause the vehicle to: acquire first driving parameters of the first vehicle and receive a first position signal sent by the second vehicle; determine a first animation effect based on the first driving parameters and generate a second animation effect based on the first position signal; determine a first display area within the first display screen where the first animation effect matches, and determine a target auxiliary display area within a plurality of auxiliary display areas where the second animation effect matches; control the first display area to play the first animation effect, and control the target auxiliary display area to play the second animation effect.

[0182] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0183] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of vehicles, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based vehicle that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0184] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0185] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described vehicle defogging method, which can solve the technical problem in related technologies that drivers cannot intuitively perceive the presence and movement trends of surrounding vehicles. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the vehicle defogging method provided in the above embodiments, and will not be repeated here.

[0186] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the vehicle defogging method described above.

[0187] The computer program product provided in this application can solve the technical problem in related technologies where drivers cannot intuitively perceive the presence and movement trends of surrounding vehicles. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the vehicle defogging method provided in the above embodiments, and will not be repeated here.

[0188] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A vehicle defogging method, characterized in that, The method includes: Acquire vehicle visual data, in-vehicle climate data, and vehicle operating condition data; Based on the vehicle visual data and preset prompt words, semantic fusion analysis is performed to obtain the matching probability of the prompt words. The preset prompt words are related to a predefined fogging scenario. Based on the matching probability of the prompt words, the scene confidence of the predefined fogging scene is obtained; If the maximum value of the scene confidence is greater than or equal to the scene confidence threshold, the scene gain is determined based on the scene confidence and the prior amplification parameter corresponding to the predefined fogging scene. Based on the in-vehicle climate data, the dew point safety margin and the short-term temperature / humidity change rate are calculated. The dew point safety margin is used to characterize the temperature difference from the dew point. The total condensation risk index is calculated based on the dew point safety margin, the short-term temperature / humidity change rate, the scene gain, and the vehicle operating condition data. If the total condensation risk index is greater than a first preset threshold, pre-defogging treatment is performed.

2. The vehicle defogging method as described in claim 1, characterized in that, The predefined fogging scenarios include outdoor scenarios, inbound scenarios, underground parking scenarios, and outbound scenarios, and the prior amplification parameters corresponding to the outdoor scenarios and underground parking scenarios are smaller than the prior amplification parameters corresponding to the inbound scenarios or the outbound scenarios.

3. The vehicle defogging method as described in claim 1, characterized in that, When the total condensation risk index is greater than a first preset threshold, the pre-defogging treatment includes: The initial defogging parameters, the in-vehicle climate data, and the vehicle operating condition data are input into the predictive control model; The predictive control model predicts short-term in-vehicle climate data and optimizes defogging parameters based on a predefined in-vehicle thermo-humid dynamics model, and solves for the target defogging parameters that meet the constraints. Pre-defogging treatment is performed based on the target defogging parameters.

4. The vehicle defogging method as described in claim 3, characterized in that, The target defogging parameters must satisfy one or more of the following constraints: maintaining windshield visibility, low energy consumption, and in-vehicle temperature.

5. The vehicle defogging method as described in claim 3, characterized in that, After performing pre-demisting treatment based on the target demisting parameters, the process further includes: Receives defogging parameters manually adjusted by the user; The weight parameters in the predictive control model are updated based on the manually adjusted defogging parameters.

6. The vehicle defogging method according to any one of claims 1-5, characterized in that, The process of obtaining the scene confidence score of the predefined fogging scene based on the matching probability of the prompt words includes: The initial confidence level of the predefined fogging scenario is calculated based on the matching probability of the prompt words; The initial confidence level is smoothed based on a preset temporal smoothing coefficient to obtain the scene confidence level of the predefined fogging scene.

7. The vehicle defogging method according to any one of claims 1-5, characterized in that, The in-vehicle climate data includes: the temperature of the inner surface of the windshield and the humidity inside the vehicle; The temperature of the inner surface of the windshield is obtained based on the vehicle's windshield temperature sensor, or estimated based on the first law of thermodynamics of the external environment temperature. The in-vehicle humidity is obtained based on the vehicle's humidity sensor, or estimated based on the vehicle volume, vehicle ventilation mass flow rate, number of occupants and metabolic level, and external environmental humidity.

8. The vehicle defogging method according to any one of claims 1-5, characterized in that, After performing pre-defogging treatment when the total condensation risk index is greater than a first preset threshold, the process further includes: Real-time acquisition of vehicle visual data, in-vehicle climate data, and vehicle operating condition data; The total condensation risk index is updated based on the vehicle visual data, the in-vehicle climate data, and the vehicle operating condition data. If the total condensation risk index is less than or equal to a second preset threshold, the pre-defogging process is stopped. The second preset threshold is less than the first preset threshold.

9. A vehicle, characterized in that, The vehicle includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the vehicle defogging method as described in any one of claims 1 to 8.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the vehicle defogging method as described in any one of claims 1 to 8.

11. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the vehicle defogging method as described in any one of claims 1 to 8.

Images